plain_tig

This model is a fine-tuned version of castorini/afriteva_v2_base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3508
• Accuracy: {'accuracy': 0.1460214446952596}

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0003
• train_batch_size: 64
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
4.8624	2.3810	100	2.5503	{'accuracy': 0.11554740406320542}
2.5725	4.7619	200	1.7106	{'accuracy': 0.13755643340857787}
1.936	7.1429	300	1.5616	{'accuracy': 0.14094243792325056}
1.804	9.5238	400	1.5510	{'accuracy': 0.14122460496613995}
1.755	11.9048	500	1.5	{'accuracy': 0.14094243792325056}
1.6713	14.2857	600	1.4747	{'accuracy': 0.14051918735891647}
1.624	16.6667	700	1.4347	{'accuracy': 0.14193002257336343}
1.5757	19.0476	800	1.4028	{'accuracy': 0.14432844243792325}
1.5407	21.4286	900	1.3813	{'accuracy': 0.14475169300225735}
1.5199	23.8095	1000	1.3686	{'accuracy': 0.1451749435665914}
1.4855	26.1905	1100	1.3569	{'accuracy': 0.14531602708803612}
1.4744	28.5714	1200	1.3508	{'accuracy': 0.1460214446952596}

Framework versions

• PEFT 0.7.1
• Transformers 4.43.3
• Pytorch 2.4.0+cu121
• Datasets 2.15.0
• Tokenizers 0.19.1